發佈會直播
講者:
鍾劍華 - 香港民意研究所副行政總裁
陳家洛 - 浸會大學政治及國際關係學系副教授、比較管治及政策研究中心總監、選舉觀察計劃成員
戴捷輝 - 香港民意研究所經理 (數據科學)
2021年10月7日香港民意研究所發佈會 – 傳媒參考資料
特別宣佈
香港民意研究計劃（香港民研）前身為香港大學民意研究計劃（港大民研）。公報內的「民研計劃」指的可以是香港民研或其前身港大民研。
公報簡要
民研計劃於昨日特首林鄭月娥發表施政報告後，即日進行調查，部分結果已於昨晚發放。是次調查除了隨機抽樣固網和手機號碼，同時加入「香港民研意見群組」中的「香港市民代表組群」(即隨機樣本組群)，並透過電話訪問或電郵邀請參與網上調查。調查的電話訪問於昨日下午約一時半開始至晚上約九時半結束，而網上調查則於下午約一時半開始至晚上約八時結束。調查共錄得936個成功個案，當中包括228個隨機抽樣固網樣本、260個隨機抽樣手機樣本、127個意見群組電話訪問樣本及321個意見群組網上調查樣本。原始數據已經按照人口比例及各抽樣架的比重加權處理，以確保數據的代表性。
調查顯示，撇除不清楚施政報告內容的被訪者後，25%表示滿意施政報告，50%不滿，滿意淨值為負25個百分點。以0至100分計，平均分為34.2分。各項數字均顯示市民對今次施政報告的評價較上年度的顯著改善。至於特首林鄭月娥，其最新評分為30.5分，在昨日發表施政報告後顯著下跌。而其支持率為19%，反對率為67%，民望淨值為負48個百分點，與發表施政報告前分別不大。此外，撇除不清楚施政報告內容的被訪者後，25%表示施政報告發表後對香港前途的信心有所增加，19%表示不變，表示減少的有50%，前途信心效應淨值為負24個百分點，數字較上年度顯著改善，但仍然屬於負面。施政報告即時調查顯示了巿民的即時反應，後續反應則有待觀察。
調查撇除意見群組樣本的實效回應比率為54.8%。在95%置信水平下，調查的百分比誤差不超過+/-4%，淨值誤差不超過+/-7%，評分誤差不超過+/-2.7。
樣本資料
| 調查日期 | ： | 6/10/2021 |
| 調查方法 | ： | (1a) 隨機抽樣固網電話訪問
(1b) 隨機抽樣手機電話訪問 (2a) 電話訪問「香港民研意見群組」中的「香港市民代表組群」 (2b) 電郵邀請「香港民研意見群組」中的「香港市民代表組群」參與網上調查 |
| 訪問對象 | ： | 18歲或以上操粵語的香港居民 |
| 成功樣本數目[1] | ： | 936 (包括228個隨機抽樣固網樣本、260個隨機抽樣手機樣本、127個意見群組電話訪問樣本及321個意見群組網上調查樣本) |
| 實效回應比率 | ： | 54.8% (撇除意見群組樣本) |
| 抽樣誤差[2] | ： | 在95%置信水平下，百分比誤差不超過+/-4%，淨值誤差不超過+/-7%，評分誤差不超過+/-2.7 |
| 加權方法 | ： | 原始數據來自4個不同的抽樣架，同時按照兩組權重以「反覆多重加權法」進行加權調整。第一組權重是政府統計處提供的人口相關統計數字，包括 a)《二零二零年年中人口數字》中全港人口年齡及性別分佈統計數字，b)《香港的女性及男性-主要統計數字》（2020年版）中的教育程度（最高就讀程度）及 c) 同一來源的經濟活動身分統計數字。第二組權重是按照各抽樣架下的預設目標樣本數目的比例調整，即固網電話隨機抽樣調查5個單位，手機電話隨機抽樣調查5個單位，意見群組樣本電話調查6個單位，及意見群組樣本網上調查4個單位。 |
[1] 數字為調查的總樣本數目，個別題目則可能只涉及次樣本。有關數字請參閱下列數表內列出的樣本數目。
[2] 此公報中所有誤差數字均以95%置信水平計算。95%置信水平，是指倘若以不同隨機樣本重複進行有關調查100次，則95次各自計算出的誤差範圍會包含人口真實數字。由於調查數字涉及抽樣誤差，傳媒引用百分比數字時，應避免使用小數點，在引用評分數字時，則可以使用一個小數點。
最新數據
以下是市民對本年度施政報告的滿意程度及過往多年的相關數字：
| 調查日期 | 樣本數目[3] | 對施政報告的評價 | |||||
| 滿意率[4] | 一半半 | 不滿率[4] | 滿意率淨值 | 平均量值[4] | 施政報告 評分 |
||
| 6/10/21 | 621 | 25+/-4%[6] | 13+/-3%[6] | 50+/-4%[6] | -25+/-7%[6] | 2.4+/-0.1[6] | 34.2+/-2.7[6] |
| 25/11/20 | 512 | 19% | 9% | 64% | -46% | 2.0 | 27.2 |
| 16/10/19 | 679 | 17%[6] | 8%[6] | 65%[6] | -47%[6] | 2.0[6] | 29.7[6] |
| 10/10/18 | 534 | 33%[6] | 24% | 34%[6] | -1%[6] | 2.9[6] | 48.5[6] |
| 11/10/17 | 526 | 48%[6] | 28%[6] | 14%[6] | 34%[6] | 3.5[6] | 62.4[6] |
| 18/1/17 | 512 | 34%[6] | 22% | 29%[6] | 5%[6] | 3.0[6] | 52.3[6] |
| 13/1/16 | 522 | 19%[6] | 23% | 39% | -20%[6] | 2.5[6] | 41.1[6] |
| 14/1/15 | 503 | 30%[6] | 24%[6] | 35% | -5%[6] | 2.8 | 49.5[6] |
| 15/1/14 | 611 | 36% | 30%[6] | 31%[6] | 5% | 3.0 | 54.1[6] |
| 16/1/13 | 759 | 36%[6] | 35% | 24%[6] | 11%[6] | 3.1 | 56.4[6] |
| 12/10/11 | 816 | 47%[6] | 32% | 18% | 28%[6] | 3.3 | 59.1 |
| 13/10/10 | 747 | 41%[6] | 33%[6] | 19%[6] | 22%[6] | 3.2 | 58.9[6] |
| 14/10/09 | 462 | 30% | 37% | 28% | 2% | 3.0 | 53.5 |
| 15/10/08 | 515 | 31%[6] | 35%[6] | 26%[6] | 4%[6] | 3.0 | 53.8[6] |
| 10/10/07 | 602 | 52%[6] | 29%[6] | 10%[6] | 42%[6] | 3.5 | 65.2[6] |
| 11/10/06 | 445 | 30%[6] | 37% | 22%[6] | 8%[6] | 3.0 | 55.8[6] |
| 12/10/05 | 377 | 48%[6] | 33% | 9%[6] | 39%[6] | 3.5 | 66.4[6] |
| 12/1/05 | 391 | 38%[6] | 30% | 20%[6] | 18%[6] | 3.2 | 56.3[6] |
| 7/1/04 | 381 | 25% | 26% | 33%[6] | -8% | 2.8 | 49.3 |
| 8/1/03[5] | 377 | 22%[6] | 29% | 27% | -5% | 2.8 | 51.6[6] |
| 10/10/01 | 433 | 29% | 33% | 28% | 1% | 3.0 | 56.7 |
| 11/10/00 | 262 | 25%[6] | 28% | 31% | -6%[6] | 2.9 | 55.2 |
| 6/10/99 | 236 | 31%[6] | 30% | 25%[6] | 6%[6] | 3.0 | 57.3 |
| 7/10/98 | 508 | 22%[6] | 35%[6] | 35%[6] | -14%[6] | 2.8 | -- |
| 8/10/97 | 534 | 45% | 30%[6] | 14%[6] | 31% | 3.4 | -- |
[3] 已撇除未聞／不知道施政報告內容而沒有作答的被訪者。民研計劃在2020年3月前彙報的次樣本數目為加權數字，2020年3月開始則以原始數字彙報。
[4] 數字採自五等量尺。平均量值是把答案按照正面程度，以1分最低5分最高量化成為1、2、3、4、5分，再求取樣本平均數值。
[5] 2003年施政報告的即時反應調查分2天進行，本表只列舉首天錄得的統計數字，以作直接比較分析之用。
[6] 該數字與上次調查結果的差異超過在95%置信水平的抽樣誤差，表示有關變化在統計學上表面成立。不過，變化在統計學上成立與否，並不等同有關變化是否有實際用途或意義，而不同調查的加權方法亦可能有所不同。
撇除不清楚施政報告內容的被訪者後，25%表示滿意施政報告，50%不滿，滿意淨值為負25個百分點，平均量值為2.4分，即整體上介乎「幾不滿」及「一半半」之間。以0至100分計，平均分為34.2分。各項數字均顯示市民對今次施政報告的評價較上年度的顯著改善。
以下是1997年至今，歷任特首在發表施政報告後的即時民望變化：
| 董建華民望 | ||||||||
| 施政報告發表日期 | 8/10/97 | 7/10/98 | 6/10/99 | 11/10/00 | 10/10/01 | 8/1/03 | 7/1/04 | 12/1/05 |
| 施政報告前評分 | 65.8 | 55.8 | 54.0 | 48.2 | 48.4 | 46.6 | 42.9 | 47.2 |
| 即時調查評分 | 66.1 | 56.1 | 54.3 | 50.7 | 50.6 | 47.3 | 44.6 | 48.4 |
| 評分變化 | +0.3 | +0.3 | +0.3 | +2.5[8] | +2.2[8] | +0.7 | +1.7[8] | +1.2 |
| 曾蔭權民望 | ||||||||
| 施政報告發表日期 | 12/10/05 | 11/10/06 | 10/10/07 | 15/10/08 | 14/10/09 | 13/10/10 | 12/10/11 | |
| 施政報告前評分 | 68.0 | 62.9 | 65.8 | 52.7 | 55.2 | 55.4 | 48.4 | |
| 即時調查評分 | 67.4 | 59.8 | 64.4 | 53.9 | 54.2 | 56.2 | 50.6 | |
| 評分變化 | -0.6 | -3.1[8] | -1.4[8] | +1.2 | -1.0 | +0.8 | +2.2[8] | |
| 施政報告前支持率淨值 | 68% | 48% | 48% | 5% | 7% | -1% | -45% | |
| 即時調查支持率淨值 | 65% | 36% | 48% | 10% | 8% | 0% | -41% | |
| 支持率淨值變化[7] | -3% | -12%[8] | -- | +5% | +1% | +1% | +4% | |
| 梁振英民望 | ||||||||
| 施政報告發表日期 | 16/1/13 | 15/1/14 | 14/1/15 | 13/1/16 | 18/1/17 | |||
| 施政報告前評分 | 48.9 | 45.6 | 40.6 | 37.5 | 41.3 | |||
| 即時調查評分 | 52.2 | 48.9 | 44.8 | 37.0 | 41.7 | |||
| 評分變化 | +3.3[8] | +3.3[8] | +4.2[8] | -0.5 | +0.4 | |||
| 施政報告前支持率淨值 | -20% | -31% | -39% | -44% | -44% | |||
| 即時調查支持率淨值 | -11% | -24% | -35% | -54% | -57% | |||
| 支持率淨值變化[7] | +9%[8] | +7%[8] | +4% | -10%[8] | -13%[8] | |||
| 林鄭月娥民望 | ||||||||
| 施政報告發表日期 | 11/10/17 | 10/10/18 | 16/10/19 | 25/11/20 | 6/10/21 | |||
| 施政報告前評分 | 59.6 | 52.3 | 22.3 | 30.8 | 33.9+/-2.0 | |||
| 即時調查評分 | 61.1 | 47.6 | 22.7 | 26.8 | 30.5+/-2.2 | |||
| 評分變化 | +1.5 | -4.7[8] | +0.3 | -4.1[8] | -3.4[8] | |||
| 施政報告前支持率淨值 | 10% | 4% | -65% | -48% | -46+/-5% | |||
| 即時調查支持率淨值 | 23% | -10% | -64% | -57% | -48+/-5% | |||
| 支持率淨值變化[7] | +13%[8] | -14%[8] | +1% | -9%[8] | -2% | |||
[7] 施政報告即時調查自2004年開始涵蓋特首支持率問題，因此沒有列入董建華施政報告調查系列。
[8] 該數字與上次調查結果的差異超過在95%置信水平的抽樣誤差，表示有關變化在統計學上表面成立。不過，變化在統計學上成立與否，並不等同有關變化是否有實際用途或意義，而不同調查的加權方法亦可能有所不同。
以下是林鄭月娥在發表施政報告前後的民望走勢：
| 調查日期 | 19-22/7/21 | 9-12/8/21 | 20-26/8/21 | 6-10/9/21 | 16-23/9/21 | 6/10/21 | 最新變化 |
| 樣本數目 | 1,000 | 1,002 | 1,003 | 1,000 | 1,036 | 936 | -- |
| 回應比率 | 48.5% | 49.4% | 52.9% | 44.2% | 44.1% | 54.8% | -- |
| 最新結果 | 結果 | 結果 | 結果 | 結果 | 結果 | 結果及 誤差 |
-- |
| 特首林鄭月娥評分 | 34.7 | 35.1 | 33.8 | 35.6 | 33.9 | 30.5+/-2.2 | -3.4[9] |
| 林鄭月娥出任特首支持率 | 20% | 20% | 20% | 24%[9] | 20%[9] | 19+/-3% | -1% |
| 林鄭月娥出任特首反對率 | 68% | 66% | 68% | 65% | 66% | 67+/-3% | +1% |
| 支持率淨值 | -48% | -46% | -48% | -41% | -46% | -48+/-5% | -2% |
[9] 該數字與上次調查結果的差異超過在95%置信水平的抽樣誤差，表示有關變化在統計學上表面成立。不過，變化在統計學上成立與否，並不等同有關變化是否有實際用途或意義，而不同調查的加權方法亦可能有所不同。
即時調查顯示，特首林鄭月娥的最新評分為30.5分，在昨日發表施政報告後顯著下跌。而其支持率為19%，反對率為67%，民望淨值為負48個百分點，與發表施政報告前分別不大。
調查就特首林鄭月娥發表其施政報告後詢問市民對香港前途信心的轉變，結果如下：
| 調查日期 | 樣本數目[10] | 對香港前途的信心 | |||
| 增加 | 不變 | 減少 | 前途信心 效應淨值 |
||
| 6/10/21 | 621 | 25+/-4%[11] | 19+/-3% | 50+/-4%[11] | -24+/-7%[11] |
| 25/11/20 | 512 | 17%[11] | 16%[11] | 63% | -46% |
| 16/10/19 | 679 | 12%[11] | 22% | 61%[11] | -49%[11] |
| 10/10/18 | 534 | 23%[11] | 25%[11] [12] | 45%[11] | -22%[11] |
| 11/10/17 | 526 | 40%[11] | 39% | 19%[11] | 21%[11] |
| 18/1/17 | 511 | 24%[11] | 36%[11] | 32%[11] | -8%[11] |
| 13/1/16 | 521 | 16%[11] | 31% | 44%[11] | -27%[11] |
| 14/1/15 | 501 | 22% | 35% | 38%[11] | -16% |
| 15/1/14 | 846 | 24%[11] | 38% | 32%[11] | -9%[11] |
| 16/1/13 | 913 | 31% | 38%[11] | 23% | 8% |
| 12/10/11 | 957 | 29% | 45% | 21% | 8%[11] |
| 13/10/10 | 914 | 31%[11] | 45% | 18%[11] | 14%[11] |
| 14/10/09 | 749 | 27%[11] | 47%[11] | 22%[11] | 5%[11] |
| 15/10/08 | 761 | 23%[11] | 38%[11] | 32%[11] | -9%[11] |
| 10/10/07 | 388 | 53%[11] | 31%[11] | 7%[11] | 46%[11] |
| 11/10/06 | 431 | 25%[11] | 51%[11] | 16%[11] | 9%[11] |
| 12/10/05 | 476 | 54%[11] | 33%[11] | 5%[11] | 49%[11] |
| 12/1/05 | 658 | 34% | 41% | 12%[11] | 22%[11] |
| 7/1/04 | 602 | 32%[11] | 40% | 16%[11] | 16%[11] |
| 8/1/03[13] | 513 | 25% | 40%[11] | 21% | 3% |
| 10/10/01 | 591 | 22% | 50%[11] | 21%[11] | 1%[11] |
| 11/10/00 | 292 | 22%[11] | 40% | 15% | 7%[11] |
| 6/10/99 | 233 | 40%[11] | 36%[11] | 16%[11] | 24%[11] |
| 7/10/98 | 505 | 21% | 52% | 22% | -1% |
[10] 已撇除未聞／不知道施政報告內容而沒有作答的被訪者。民研計劃在2020年3月前彙報的次樣本數目為加權數字，2020年3月開始則以原始數字彙報。
[11] 該數字與上次調查結果的差異超過在95%置信水平的抽樣誤差，表示有關變化在統計學上表面成立。不過，變化在統計學上成立與否，並不等同有關變化是否有實際用途或意義，而不同調查的加權方法亦可能有所不同。
[12] 該數字與上次調查結果的差異超過在95%置信水平的抽樣誤差，是由於加權方法改變。如果以舊有加權方法處理數據，則變化並未超過抽樣誤差。
[13] 2003年施政報告的即時反應調查分2天進行，本表只列舉首天錄得的統計數字，以作直接比較分析之用。
調查顯示，撇除不清楚施政報告內容的被訪者後，25%表示施政報告發表後對香港前途的信心有所增加，19%表示不變，表示減少的有50%，前途信心效應淨值為負24個百分點，數字較上年度顯著改善，但仍然傾向負面。
數據分析
最新的施政報告即時調查顯示，撇除不清楚施政報告內容的被訪者後，25%表示滿意施政報告，50%不滿，滿意淨值為負25個百分點。以0至100分計，平均分為34.2分。各項數字均顯示市民對今次施政報告的評價較上年度的顯著改善。
至於特首林鄭月娥，其最新評分為30.5分，在昨日發表施政報告後顯著下跌。而其支持率為19%，反對率為67%，民望淨值為負48個百分點，與發表施政報告前分別不大。
此外，撇除不清楚施政報告內容的被訪者後，25%表示施政報告發表後對香港前途的信心有所增加，19%表示不變，表示減少的有50%，前途信心效應淨值為負24個百分點，數字較上年度顯著改善，但仍然傾向負面。
施政報告即時調查顯示了巿民的即時反應，後續反應則有待觀察。
Oct 07, 2021
Hong Kong Public Opinion Research Institute Press Conference – Press Materials
Press Conference Live
Speakers:
Kim-Wah Chung - Deputy CEO, HKPORI
Kenneth Chan - Associate Professor, Department of Government & International Studies, HKBU; Director, Comparative Governance & Policy Research Centre; Member, Election Observation Project
Edward Tai - Manager (Data Science), HKPOR
Oct 7, 2021
Hong Kong Public Opinion Research Institute Press Conference – Press Materials
Special Announcement
The predecessor of Hong Kong Public Opinion Program (HKPOP) was The Public Opinion Programme at The University of Hong Kong (HKUPOP). “POP” in this release can refer to HKPOP or its predecessor HKUPOP.
Abstract
After Chief Executive Carrie Lam delivered the Policy Address yesterday, POP conducted an instant survey on the same day and released part of the findings last night. Apart from random landline and mobile numbers, this survey also included samples from our “Hong Kong People Representative Panel” (i.e., a panel comprising randomly recruited samples) within “HKPOP Panel”, interviewed by telephone or invited through email to complete an online survey. Our telephone survey began at around 1:30pm till around 9:30pm, while our online survey started at around 1:30pm and ended at around 8pm yesterday. A total of 936 successful cases were collected, including 228 random landline samples, 260 random mobile samples, 127 panel telephone survey samples and 321 panel online survey samples. The raw data have been weighted by population statistics and proportions of different sampling frames to ensure data representativeness.
Our survey shows that after excluding those respondents who said they did not have any knowledge of the Policy Address, 25% said they were satisfied with it, 50% were dissatisfied, giving a net satisfaction rate of negative 25 percentage points. On a scale of 0-100, the average rating is 34.2 marks. Various figures show that people’s appraisal of this Policy Address has significantly improved compared to last year. As for CE Carrie Lam, her latest support rating is 30.5 marks, which has dropped significantly after she delivered her Policy Address yesterday. Meanwhile, her approval rate stands at 19% and disapproval stands at 67%, giving a net approval rate of negative 48 percentage points, which has not changed much compared to before the Policy Address was delivered. Moreover, after excluding those who did not have any knowledge of the Policy Address, 25% said their confidence in the future of Hong Kong had increased, 19% said no change, while 50% said their confidence had decreased, giving a net effect of negative 24 percentage points on people’s confidence. The figure has significantly improved compared to last year, but it is still negative. The instant survey describes people’s instant reaction toward the Policy Address. Their reactions later remain to be seen.
The effective response rate of the survey excluding panel samples is 54.8%. The maximum sampling error of percentages is +/-4%, that of net values is +/-7% and that of ratings is +/-2.7 at 95% confidence level.
Contact Information
| Date of survey | : | 6/10/2021 |
| Survey method | : | (1a) Random landline telephone survey
(1b) Random mobile telephone survey (2a) Telephone survey targeting “Hong Kong People Representative Panel” within “HKPOP Panel” (2b) Online survey with email invitation targeting “Hong Kong People Representative Panel” within “HKPOP Panel” |
| Target population | : | Cantonese-speaking Hong Kong residents aged 18 or above |
| Sample size[1] | : | 936 (including 228 random landline samples, 260 random mobile samples, 127 panel telephone survey samples and 321 panel online survey samples) |
| Effective response rate | : | 54.8% (excluding panel samples) |
| Sampling error[2] | : | Sampling error of percentages not more than +/-4%, that of net values not more than +/-7% and that of ratings not more than +/-2.7 at 95% conf. level |
| Weighting method | : | The raw data comes from 4 different sampling frames. It is rim-weighted by two sets of weighting factors simultaneously. The first set of weighting factors comprises population figures provided by the Census and Statistics Department, they include (a) the gender-age distribution of the Hong Kong population from “Mid-year population for 2020”, (b) educational attainment (highest level attended) distribution from “Women and Men in Hong Kong - Key Statistics (2020 Edition)”, and (c) economic activity status distribution from the last source. The second set of weighting factors is adjusted based on the relative target sample size of different sub-sampling frames, namely, random telephone survey using landline numbers set at 5 units, random telephone survey using mobile numbers set at 5 units, telephone survey of randomly pre-selected panel members set at 6 units, and online survey of randomly pre-selected panel members set at 4 units. |
[1] This figure is the total sample size of the survey. Some questions may only involve a subsample, the size of which can be found in the tables below.
[2] All error figures in this release are calculated at 95% confidence level. “95% confidence level” means that if we were to repeat a certain survey 100 times with different random samples, we would expect 95 times having the population parameter within the respective error margins calculated. Because of sampling errors, when quoting percentages, journalists should refrain from reporting decimal places, whereas one decimal place can be used when quoting rating figures.
Latest Figures
People’s satisfaction figures with this year’s Policy Address are summarized below together with the previous findings:
| Date of survey | Sample size[3] | Appraisal of Policy Address | |||||
| Satisfaction rate[4] | Half-half | Dissatisfaction rate[4] | Net satisfaction rate | Mean value[4] | Rating of Policy Address |
||
| 6/10/21 | 621 | 25+/-4%[6] | 13+/-3%[6] | 50+/-4%[6] | -25+/-7%[6] | 2.4+/-0.1[6] | 34.2+/-2.7[6] |
| 25/11/20 | 512 | 19% | 9% | 64% | -46% | 2.0 | 27.2 |
| 16/10/19 | 679 | 17%[6] | 8%[6] | 65%[6] | -47%[6] | 2.0[6] | 29.7[6] |
| 10/10/18 | 534 | 33%[6] | 24% | 34%[6] | -1%[6] | 2.9[6] | 48.5[6] |
| 11/10/17 | 526 | 48%[6] | 28%[6] | 14%[6] | 34%[6] | 3.5[6] | 62.4[6] |
| 18/1/17 | 512 | 34%[6] | 22% | 29%[6] | 5%[6] | 3.0[6] | 52.3[6] |
| 13/1/16 | 522 | 19%[6] | 23% | 39% | -20%[6] | 2.5[6] | 41.1[6] |
| 14/1/15 | 503 | 30%[6] | 24%[6] | 35% | -5%[6] | 2.8 | 49.5[6] |
| 15/1/14 | 611 | 36% | 30%[6] | 31%[6] | 5% | 3.0 | 54.1[6] |
| 16/1/13 | 759 | 36%[6] | 35% | 24%[6] | 11%[6] | 3.1 | 56.4[6] |
| 12/10/11 | 816 | 47%[6] | 32% | 18% | 28%[6] | 3.3 | 59.1 |
| 13/10/10 | 747 | 41%[6] | 33%[6] | 19%[6] | 22%[6] | 3.2 | 58.9[6] |
| 14/10/09 | 462 | 30% | 37% | 28% | 2% | 3.0 | 53.5 |
| 15/10/08 | 515 | 31%[6] | 35%[6] | 26%[6] | 4%[6] | 3.0 | 53.8[6] |
| 10/10/07 | 602 | 52%[6] | 29%[6] | 10%[6] | 42%[6] | 3.5 | 65.2[6] |
| 11/10/06 | 445 | 30%[6] | 37% | 22%[6] | 8%[6] | 3.0 | 55.8[6] |
| 12/10/05 | 377 | 48%[6] | 33% | 9%[6] | 39%[6] | 3.5 | 66.4[6] |
| 12/1/05 | 391 | 38%[6] | 30% | 20%[6] | 18%[6] | 3.2 | 56.3[6] |
| 7/1/04 | 381 | 25% | 26% | 33%[6] | -8% | 2.8 | 49.3 |
| 8/1/03[5] | 377 | 22%[6] | 29% | 27% | -5% | 2.8 | 51.6[6] |
| 10/10/01 | 433 | 29% | 33% | 28% | 1% | 3.0 | 56.7 |
| 11/10/00 | 262 | 25%[6] | 28% | 31% | -6%[6] | 2.9 | 55.2 |
| 6/10/99 | 236 | 31%[6] | 30% | 25%[6] | 6%[6] | 3.0 | 57.3 |
| 7/10/98 | 508 | 22%[6] | 35%[6] | 35%[6] | -14%[6] | 2.8 | -- |
| 8/10/97 | 534 | 45% | 30%[6] | 14%[6] | 31% | 3.4 | -- |
[3] Respondents who did not answer this question because they had not heard of / did not have any knowledge of the Policy Address have been excluded. Before March 2020, weighted count was used to report subsample size. Starting from March 2020, raw count was used instead.
[4] Collapsed from a 5-point scale. The mean value is calculated by quantifying all individual responses into 1, 2, 3, 4, 5 marks according to their degree of positive level, where 1 is the lowest and 5 the highest, and then calculate the sample mean.
[5] The 2003 Policy Address instant survey was conducted for two days. Only figures registered in the first day of fieldwork are listed in this table for direct comparison and analysis.
[6] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level, meaning that the change is statistically significant prima facie. However, whether the difference is statistically significant is not the same as whether they are practically useful or meaningful, and different weighting methods could have been applied in different surveys.
After excluding those respondents who said they did not have any knowledge of the Policy Address, 25% said they were satisfied with it, 50% were dissatisfied, giving a net satisfaction rate of negative 25 percentage points. The mean score is 2.4, meaning between “quite dissatisfied” and “half-half” in general. On a scale of 0-100, the average rating is 34.2 marks. Various figures show that people’s appraisal of this Policy Address has significantly improved compared to last year.
Figures on various Chief Executives’ popularity before and after the Policy Address Speech since 1997 are summarized as follows:
| Popularity of Tung Chee-hwa | ||||||||
| Date of PA Speech | 8/10/97 | 7/10/98 | 6/10/99 | 11/10/00 | 10/10/01 | 8/1/03 | 7/1/04 | 12/1/05 |
| Rating before the PA | 65.8 | 55.8 | 54.0 | 48.2 | 48.4 | 46.6 | 42.9 | 47.2 |
| Rating at instant survey | 66.1 | 56.1 | 54.3 | 50.7 | 50.6 | 47.3 | 44.6 | 48.4 |
| Change in rating | +0.3 | +0.3 | +0.3 | +2.5[8] | +2.2[8] | +0.7 | +1.7[8] | +1.2 |
| Popularity of Donald Tsang | ||||||||
| Date of Policy Address Speech | 12/10/05 | 11/10/06 | 10/10/07 | 15/10/08 | 14/10/09 | 13/10/10 | 12/10/11 | |
| Rating before the PA | 68.0 | 62.9 | 65.8 | 52.7 | 55.2 | 55.4 | 48.4 | |
| Rating at instant survey | 67.4 | 59.8 | 64.4 | 53.9 | 54.2 | 56.2 | 50.6 | |
| Change in rating | -0.6 | -3.1[8] | -1.4[8] | +1.2 | -1.0 | +0.8 | +2.2[8] | |
| Net approval rate before the PA | 68% | 48% | 48% | 5% | 7% | -1% | -45% | |
| Net approval rate at instant survey | 65% | 36% | 48% | 10% | 8% | 0% | -41% | |
| Change in net approval rate[7] | -3% | -12%[8] | -- | +5% | +1% | +1% | +4% | |
| Popularity of CY Leung | ||||||||
| Date of Policy Address Speech | 16/1/13 | 15/1/14 | 14/1/15 | 13/1/16 | 18/1/17 | |||
| Rating before the PA | 48.9 | 45.6 | 40.6 | 37.5 | 41.3 | |||
| Rating at instant survey | 52.2 | 48.9 | 44.8 | 37.0 | 41.7 | |||
| Change in rating | +3.3[8] | +3.3[8] | +4.2[8] | -0.5 | +0.4 | |||
| Net approval rate before the PA | -20% | -31% | -39% | -44% | -44% | |||
| Net approval rate at instant survey | -11% | -24% | -35% | -54% | -57% | |||
| Change in net approval rate[7] | +9%[8] | +7%[8] | +4% | -10%[8] | -13%[8] | |||
| Popularity of Carrie Lam | ||||||||
| Date of Policy Address Speech | 11/10/17 | 10/10/18 | 16/10/19 | 25/11/20 | 6/10/21 | |||
| Rating before the PA | 59.6 | 52.3 | 22.3 | 30.8 | 33.9+/-2.0 | |||
| Rating at instant survey | 61.1 | 47.6 | 22.7 | 26.8 | 30.5+/-2.2 | |||
| Change in rating | +1.5 | -4.7[8] | +0.3 | -4.1[8] | -3.4[8] | |||
| Net approval rate before the PA | 10% | 4% | -65% | -48% | -46+/-5% | |||
| Net approval rate at instant survey | 23% | -10% | -64% | -57% | -48+/-5% | |||
| Change in net approval rate[7] | +13%[8] | -14%[8] | +1% | -9%[8] | -2% | |||
[7] Instant surveys on Policy Address included CE’s approval rate since 2004, so it is not listed under Tung’s series.
[8] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level, meaning that the change is statistically significant prima facie. However, whether the difference is statistically significant is not the same as whether they are practically useful or meaningful, and different weighting methods could have been applied in different surveys.
Recent figures on Carrie Lam’s popularity before and after the Policy Address speech are as follows:
| Date of survey | 19-22/7/21 | 9-12/8/21 | 20-26/8/21 | 6-10/9/21 | 16-23/9/21 | 6/10/21 | Latest change |
| Sample size | 1,000 | 1,002 | 1,003 | 1,000 | 1,036 | 936 | -- |
| Response rate | 48.5% | 49.4% | 52.9% | 44.2% | 44.1% | 54.8% | -- |
| Latest findings | Finding | Finding | Finding | Finding | Finding | Finding & error | -- |
| Rating of CE Carrie Lam | 34.7 | 35.1 | 33.8 | 35.6 | 33.9 | 30.5+/-2.2 | -3.4[9] |
| Vote of confidence in CE Carrie Lam |
20% | 20% | 20% | 24%[9] | 20%[9] | 19+/-3% | -1% |
| Vote of no confidence in CE Carrie Lam |
68% | 66% | 68% | 65% | 66% | 67+/-3% | +1% |
| Net approval rate | -48% | -46% | -48% | -41% | -46% | -48+/-5% | -2% |
[9] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level, meaning that the change is statistically significant prima facie. However, whether the difference is statistically significant is not the same as whether they are practically useful or meaningful, and different weighting methods could have been applied in different surveys.
Instant survey shows that CE Carrie Lam’s latest support rating is 30.5 marks, which has dropped significantly after she delivered her Policy Address yesterday. Meanwhile, her approval rate stands at 19% and disapproval stands at 67%, giving a net approval rate of negative 48 percentage points, which has not changed much compared to before the Policy Address was delivered.
The survey also gauged the change of people’s confidence in the future of Hong Kong after CE Carrie Lam delivered her Policy Address. Results are as follows:
| Date of survey |
Sample size[10] |
Confidence in the future of Hong Kong | |||
| Increased | Unchanged | Decreased | Net effect on confidence | ||
| 6/10/21 | 621 | 25+/-4%[11] | 19+/-3% | 50+/-4%[11] | -24+/-7%[11] |
| 25/11/20 | 512 | 17%[11] | 16%[11] | 63% | -46% |
| 16/10/19 | 679 | 12%[11] | 22% | 61%[11] | -49%[11] |
| 10/10/18 | 534 | 23%[11] | 25%[11] [12] | 45%[11] | -22%[11] |
| 11/10/17 | 526 | 40%[11] | 39% | 19%[11] | 21%[11] |
| 18/1/17 | 511 | 24%[11] | 36%[11] | 32%[11] | -8%[11] |
| 13/1/16 | 521 | 16%[11] | 31% | 44%[11] | -27%[11] |
| 14/1/15 | 501 | 22% | 35% | 38%[11] | -16% |
| 15/1/14 | 846 | 24%[11] | 38% | 32%[11] | -9%[11] |
| 16/1/13 | 913 | 31% | 38%[11] | 23% | 8% |
| 12/10/11 | 957 | 29% | 45% | 21% | 8%[11] |
| 13/10/10 | 914 | 31%[11] | 45% | 18%[11] | 14%[11] |
| 14/10/09 | 749 | 27%[11] | 47%[11] | 22%[11] | 5%[11] |
| 15/10/08 | 761 | 23%[11] | 38%[11] | 32%[11] | -9%[11] |
| 10/10/07 | 388 | 53%[11] | 31%[11] | 7%[11] | 46%[11] |
| 11/10/06 | 431 | 25%[11] | 51%[11] | 16%[11] | 9%[11] |
| 12/10/05 | 476 | 54%[11] | 33%[11] | 5%[11] | 49%[11] |
| 12/1/05 | 658 | 34% | 41% | 12%[11] | 22%[11] |
| 7/1/04 | 602 | 32%[11] | 40% | 16%[11] | 16%[11] |
| 8/1/03[13] | 513 | 25% | 40%[11] | 21% | 3% |
| 10/10/01 | 591 | 22% | 50%[11] | 21%[11] | 1%[11] |
| 11/10/00 | 292 | 22%[11] | 40% | 15% | 7%[11] |
| 6/10/99 | 233 | 40%[11] | 36%[11] | 16%[11] | 24%[11] |
| 7/10/98 | 505 | 21% | 52% | 22% | -1% |
[10] Respondents who did not answer this question because they had not heard of / did not have any knowledge of the Policy Address have been excluded. Before March 2020, weighted count was used to report subsample size. Starting from March 2020, raw count was used instead.
[11] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level, meaning that the change is statistically significant prima facie. However, whether the difference is statistically significant is not the same as whether they are practically useful or meaningful, and different weighting methods could have been applied in different surveys.
[12] The difference between the figure and the result from the previous survey has gone beyond the sampling error at 95% confidence level because of a change in the weighting method. If the previous weighting method was used, the changes would not have gone beyond the sampling errors.
[13] The 2003 Policy Address instant survey was conducted for two days. Only figures registered in the first day of fieldwork are listed in this table for direct comparison and analysis.
Results show that after excluding those who did not have any knowledge of the Policy Address, 25% said their confidence in the future of Hong Kong had increased, 19% said no change, while 50% said their confidence had decreased, giving a net effect of negative 24 percentage points on people’s confidence. The figure has significantly improved compared to last year, but it is still negative.
Data Analysis
Our latest Policy Address instant survey shows that after excluding those respondents who said they did not have any knowledge of the Policy Address, 25% said they were satisfied with it, 50% were dissatisfied, giving a net satisfaction rate of negative 25 percentage points. On a scale of 0-100, the average rating is 34.2 marks. Various figures show that people’s appraisal of this Policy Address has significantly improved compared to last year.
As for CE Carrie Lam, her latest support rating is 30.5 marks, which has dropped significantly after she delivered her Policy Address yesterday. Meanwhile, her approval rate stands at 19% and disapproval stands at 67%, giving a net approval rate of negative 48 percentage points, which has not changed much compared to before the Policy Address was delivered.
Moreover, after excluding those who did not have any knowledge of the Policy Address, 25% said their confidence in the future of Hong Kong had increased, 19% said no change, while 50% said their confidence had decreased, giving a net effect of negative 24 percentage points on people’s confidence. The figure has significantly improved compared to last year, but it is still negative.
The instant survey describes people’s instant reaction toward the Policy Address. Their reactions later remain to be seen.
